A new scheme is presented for the parameterization of pharmaceuticals in the context of the perturbed-chain statistical associating fluid theory (PC-SAFT). The pharmaceutical parameters are fitted to experimental solubility values in three different solvents, i.e., a hydrophilic, a polar, and a hydrophobic solvent. Subsequently, the solubility of the pharmaceuticals in other solvents is predicted without the need for additional interaction parameters. The approach is presented for six commonly used pharmaceutical compounds (paracetamol. naproxen, ibuprofen, flurbiprofen, ketoprofen, and lovastatin) in numerous pure and mixed solvents, and in different temperatures. (C) 2010 Elsevier B.V. All rights reserved.